The Shifting Party Coalitions--From
the 1930s to the 1970s

by Everett Carll Ladd

A political transformation of unusual speed and scope took place in the United States during the 1930s. Its most momentous feature was the emergence of a new reigning public philosophyNew Deal liberalism. But there appeared in response as well a new party system, defined by the Democrats' assuming majority status and by dramatic shifts in the composition of the partisan coalitions. By the end of the 1930s a distinctive set of voting patterns and social group attachments had been firmly established. The "New Deal coalitions" had become a prominent and familiar part of the American political landscape.

Since about 1960, though, the New Deal coalitions have been unraveling in the face of complex social change. The United States has entered upon a new sociopolitical setting most aptly described as "postindustrial"involving such conditions as a dramatic increase in national wealth, advanced technological development, the central importance of knowledge, the elaboration of national electronic communications processes, new occupational structures, and with them new life-styles and expectations, which is to say, new social classes and new centers of power.

The social and political world with which the parties must deal has changed so markedly from what it was in the age of Franklin Roosevelt that the party system could not help but be altered. A different mix of policy issues has been thrust upon the political agenda. Lines of social conflict have shifted. New interest groups have appeared and old groups have found themselves with new interests. And as the Democratic and Republican parties have grappled with this changing en-vironment, they have acted in ways which, inevitably, have increased the impact of the broad external social changes on the composition of their respective coalitions.

Today the party alliances and voting patterns which gave a distinctive stamp to the New Deal era have vanished. Some features of the past, of course, are always embedded in the present. We insist that the New Deal coalitions have left the scene, not because nothing of them remains, but because the alliances, issues, and cleavages which distinguished them are no longer those which demarcate the Democratic and Republican coalitions.

The argument of this chapter with regard to the shifting partisan alliances invokes a common observation about social change. At some point, the differences from earlier periods cannot properly be described simply as "more." The quantitative progression produces qualitative change. The analogy of a small snowball at the top of an inclination gradual at the start and becoming ever steeper is not inappropriate. The ball of snow begins to roll, slowly at first--and with its small mass, it grows but slowly; but as the mass enlarges and the inclination becomes steeper, the growth in size becomes extremely rapid. As it approaches the base of the hill, the innocent little snowball has become a fast-moving boulder of snow. Both the little ball and the boulder have something in common, but a person standing in their respective paths could not fail to detect a real difference. At what point did the little snowball become a boulder ? At what point did the New Deal coalitions vanish, to be replaced by new alliances? There is some ambiguity, but it need not concern us here. Now, at the beginning of the 1980s, the decisive transformation of the American party coalitions from their New Deal form has surely occurred.

During the New Deal era, each of the two main parties had its reasonably secure bases among certain social groups. The Republicans were the party of the Northeast, of business, of the middle classes, and of white Protestants, while the Democrats enjoyed a clear majority among the working classes, organized labor, Catholics, and the South. And these sources of respective party strength could be counted upon at all levels of the election process--in voting for president, Congress, lesser offices, and in underlying party identification. After 1960, however, this condition of relatively secure party bailiwicks defined by ethnicity, class, and region began breaking down rapidly. The Democrats emerged almost everywhere outside the presidential arena as the "everyone" party. The depiction is not intended literally, of course. Rather, it is meant to describe a novel situation in which one party had more, at least nominal, adherents than its opposition among virtually all relevant groups. Since 1960 few social collectivities have given the Republicans regular pluralities either in party identification or in the sweep of subpresidential voting.

An examination of the party identification of various age strata, social groups, and ideological clusters in 1976 shows how far the Democratic surge had carried by that date. The Democrats, for example, were well ahead of the GOP in every age cohort, from the youngest segments of the electorate to the oldest, and their margin was remarkably uniform. The proportion of self-described independents did rise steadily, with movement from the oldest to the youngest voters. The latter had had less time to establish regular preferences. And, of course, college-trained people, more inclined to see themselves as "independent," were more heavily represented in the younger cohorts. Still, there was a big and rather even Democratic lead over the Republicans, one that extended from those people whose earliest political memories go back to the 1910s and 1920s to those who first saw U.S. politics in the years of Kennedy, Johnson, and Nixon.

Tables 1 and 2 testify to the breadth of the Democrats' appeal, as compared to that of the Republicans, by the mid-1970s. Wage workers were less Republican than businessmen and executives, but a plurality of even the latter identified with the Democrats. Less than one-third of the business/professional stratum claimed attachment to the GOP. All educational groups showed a Democratic margin. So did all income cohorts up to the very prosperous. The Democrats led the Republicans in every region, among all religious groups, among virtually all ethnic groups. People from wealthy family backgrounds preferred the Democrats by a 2-to-1 margin.

The Democratic lead extended not only to most of the demographic units, but to the principal ideological groups as well. Survey work by Daniel Yankelovich, for example, showed that the Democrats far outdistanced Republicans among voters who thought of themselves as liberals and moderates, and that they even had a comfortable edge among self-described conservatives. By 1976 there were only a few Republican bailiwicks left.

In defining them, the prominent role of region and ethnicity is striking. White Protestants in the Northeast remained strongly Republican, continuing a regional/ethnic tradition that reaches far back into U.S. history. And no group was more decisively Republican than the "Yankees," if the term is taken to mean white Protestants of British stock residing in the Northeastern states.

Tables 1 and 2 show only party identification. But data on congressional voting reveal the same pattern. The Democrats had surpassed their GOP opponents across most of the social spectrum. The middle classes gave higher proportions of their vote to Republican congressional candidates than did the working classes, but both groups consistently produced large Democratic pluralities.


Over the late 1960s and early 1970s, the United States developed a two-tier system of voting, with one pattern applying at the presidential level and another almost everywhere else. This stood in contrast to the experience of the New Deal era, when presidential voting and that for the lower offices went more or less in tandem. The Democrats came to control by massive margins the Congress, the state legislatures, the governorships, and various lesser offices reflecting the breadth and inclusiveness of their coalition as described above. But, at the same time, this big majority party had extraordinary difficulties with the presidency . A Democratic presidential nominee received an absolute majority of the popular vote only twice from 1948 to 1976, and only once was there a decisive Democratic victory--Johnson's in 1964. Adding up all of the presidential votes cast from 1948 through 1976, one sees the Republicans actually leading the Democrats, 271 million to 256 million. It is not surprising, considering this, that the presidential coalitions came to differ very substantially from those formed in other contests.

One reason for the sharp and persisting disparity between the Democrats' uncertain performance in presidential elections and their domination elsewhere is that voting for the presidency came to have less and less to do with political party. The parties have been weakened organizationally since 1960, and the American electorate--better educated, more leisured, with more sources of political information, and hence more confident of its ability to judge candidates and programs without reference to their partisan origins--is less strongly attached to the parties and more inclined to vote in dependently than ever before. The emergence of television as the principal source of information about candidates, especially at the national level where media resources are so great, further transformed contests for the presidency (and a few other highly visible offices in large media-rich states such as New York and California), again reducing the party role. By contrast, in the less visible, less media-attended offices where candidates are not so well known, party is much more influential. Since the Democratic party had more adherents and was more highly regarded than the GOP, the more party emphasis the office received, the better the Democrats were likely to fare.

But it is the special symbolic importance of the presidency that led it to present particular difficulties for the Democrats. The intense public visibility of the office almost requires that various groups and interests attend closely to the style and emphasis of presidential leadership. Presidential contests thus became the principal arena in which were fought out newly emerging policy or ideological differences.

The late 1960s and early 1970s in the United States were a period of extraordinary social change and turbulence. The liberalism that shaped the New Deal state was incorporated into a national consensus, but at the same time a "New Liberalism" emerged--and it was hardly consensual. Comprising such positions as support for new standards of personal morality, cultural values, and life-styles, a questioning of the merits of economic growth on the grounds that growth threatens the "quality of life," and a redefinition of the demands of equality in the direction of "equality of result,"; the New Liberalism was fraught with danger for candidates who espoused, it, simply because a large majority of the population did not share its enthusiasm for social and cultural change. Dissatisfaction with New Liberalism was especially strong within the white working class. Wheras policy innovation in the 1930s often involved efforts by the working class to strengthen its position vis-a-vis business, in the 1960s and 1970s the New Liberals' projects for social change imposed some significant costs and risks on broad sectors of the working and lower-middle classes, who did not hesitate to make their unhappiness known.

The New Liberalism was not, however, without its support, especially among more affluent sectors of the society--even though by 1980 it was decisively out of vogue. The upper-middle class, altered by the infusion of a large professional/managerial, public-sector cohort, had been detached from traditional business concerns. Its perspectives seemed to be shaped far more by the universities and the intellectual community, and many of its members had come to share in some of the critical, change--demanding orientations that had long been associated with intellectual life. These upper strata also had a natural concern about life-styles and cultural change--areas in which the higher- status groups have always been more receptive to change than the traditional middle and working classes. And because they were more secure in their position--less threatened by many contemporary efforts at social change, typically residing some distance from the "front lines"the upper strata came easily to a more change-supportive posture in such critical areas as race relations.

The Democrats became the primary party of proponents of both liberalisms. The Democratic coalition was thus divided, and over the late 1960s and 1970s the tendency to call attention to this division was far greater at the presidential level than in other contests. The New Liberalism impacted too strongly on the national stage; it was too divisive for the Democrats to be able to package, with any degree of regularity, a ticket which left both New Liberals and Old Liberals comfortably allied. Surely the most dramatic instance of this schism was the 1972 presidential race, in which George McGovern took up the New Liberal banner and was buried electorally by a never especially popular Republican. At the congressional level and elsewhere, however, the party was able to stress the more consensual Old Liberal themes and had little difficulty keeping its massive, heterogeneous coalition together.

Still, Carter failed to win a majority of the vote among Southern whites. Every "Yankee" Democratic nominee from Samuel Tilden through Adlai Stevenson won majority backing among white Southerners--and most did so by overwhelming margins. Southern white Protestants were a loyal and numerically substantial component of the Roosevelt presidential coalition. They gave a higher proportion of their ballots to FDR than any other large, politically relevant, social collectivity. But in 1976 the group was 5 percentage points less Democratic than the national electorate. [2] On only four occasions from the Civil War to 1976 did a Democratic presidential nominee lose majority support in the white Protestant South. And these were four presidential elections between 1960 and 197--all of the elections in this span, that is, except that of 1964. (Again in 1980, Democrat Carter lost the white South.) In many ways, the 1976 results provided even more dramatic proof of the scope and permanence of the presidential realignment than did the McGovern-Nixon contest. If a centrist white Southerner--starting with an extraordinarily high margin of support in his native Deep South state, contesting against the Republican party that had been mauled by the most dramatic political scandal in the country's history and burdened with a poorly performing economy, pitted against a Republican nominee who was the second choice of his own party in the region and was scarcely the most forceful or charismatic of contenders--could not win majority backing in the white Protestant South, what Democratic nominee ever can?


The Democrats' position nationally as the principal partisan instrument for the New Liberalism brought them a long-term disaffection among other groups that were important parts of the old New Deal majority. Manual workers, big-city dwellers, and Catholics, along with Southern whites, consistently gave much higher majorities during the Roosevelt era to the Democratic presidential nominee than did the public at large. The country was Democratic, and these groups were notably more Democratic than the country. By the elections of 1968 and 1972, however, this Democratic margin had vanished. Blue-collar whites, for instance12 percent more Democratic than the populace generally in the 1940 election, and again, 12 percent more Democratic in 1948--by 1968 gave the Democrats a proportion of their ballots only 3 points higher than the entire electorate, and by 1972 were actually 4 points less Democratic than all voters. There was a comparable decline in the relative margin of Democratic support among such overlapping groups as urban Catholic voters and big-city, working-class whites outside the South (figure 2).

The 1976 vote distribution showed a good bit of similarity to those of other recent presidential elections. The Democrats in that year did not regain the large margins within the white working classes which contributed so importantly to their ascendancy in the New Deal era. Urban white Catholics, and blue-collar workers other than those who are blacks and Hispanics, like Southern white Protestants, were somewhat more Democratic in 1976 than they had been in the McGovern-Nixon election, but their electoral standing remained consonant with that of the last decade and a half. There was no return to the Roosevelt-era high Democratic performance.

So the "old New Deal Democratic coalition" was not "put back together one more time" in 1976. Jimmy Carter surely received substantial backing from groups that were the building blocks of the Rooseveltian alignment, and he improved markedly on McGovern's electoral performance, but the Carter coalition had a quite different base than FDR's. Well before the party's 1980 debacle, the character and makeup of the Democrats' presidential coalition had decisively changed.


Class lines were unusually salient in the electoral behavior of the New Deal and immediate post-New Deal years, with groups of lower socioeconomic status markedly more Democratic than the high-status cohorts. This resulted in a relationship involving social class and ideology. That is, the lower-status groups were rather consistently more supportive of liberal programs and policies in the New Deal context than were their high-status counterparts; because the Democrats were the party of New Deal liberalism, they drew their backing disproportionately from among people of lower socioeconomic standing.

Over the late 1960s and early 1970s, however, a pronounced inversion of the New Deal class/ideology relationship occurred--with the result that high-status groups in the United States became more liberal, given the current construction of that category, than the middle-to-lowerstatus groups across a wide array of issues. This inversion, it must be emphasized, developed around class and ideology, and it extended to class and party only in those cases and to the extent that one party was clearly associated with the contemporary extensions of liberalism. The national Democratic party, far more than its Republican opposition, became home in the 1960s and 1970s to the New Liberalism, but the extent of the latter association varied with the individual candidates put forward in each election.

In the presidential elections of 1968 and 1972, the issues of the New Liberalism were unusually prominent, and the Democrats suffered a marked falling off in their relative standing among lower-status groups as compared with their support among high- status cohorts. There was something approaching an actual inversion of the familiar New Deal class/party alignment. In many instances in 1968 and 1972, groups at the top were more Democratic in their presidential voting than those at the bottom.

Table 3 reviews the classic pattern of class voting as it persisted throughout the presidential elections of the New Deal era and into the early 1960s. What had been well established as the traditional configuration held neatly for the several sets of groups represented in this table, and indeed, for all the various socioeconomic groupings which can be located with survey data. Thus, only 30 percent of all white Americans of high socioeconomic status voted for Democratic nominee Harry Truman in 1948, compared with 43 percent for Truman among middle-status whites and 57 percent among low-status white voters. In the 1960 Kennedy-Nixon election the same relationship can be seen, with Kennedy backed by just 38 percent of the high-status group, by 53 percent of the middle-status, and by 61 percent of those of low socioeconomic standing.

By 1968, however, the relationship had changed markedly. Humphrey was supported that year by 50 percent of high status whites under thirty years of age, but by only 39 percent of their middle-status age mates and by just 32 percent of young, low-status electors. The newly emergent conformation was even clearer in 1972, when the somewhat distorting factor of the Wallace candidacy was removed. Among whites--for blacks constitute a deviating case of voters who are disproportionately in the lower socioeconomic strata but overwhelmingly Democratic--those with college training were more Democratic than those who had not attended college; persons in the professional and managerial stratum were more Democratic than the semi-skilled and unskilled work force; and so on. McGovern was backed by 45 percent of the college-educated young, but by only 30 percent of their age mates who had not entered the groves of academe. Comparing 1948 and 1972, we see a reversal of quite extraordinary proportions.

We may conclude, without need of extensive elaboration, that the McGovern candidacy of 1972 distinguished itself far more than the 1976 Carter candidacy by a commitment to the style, programs, and constituencies of the New Liberalism. The tendencies to an electoral inversion paralleling those of class and ideology, then, should have been more pronounced in 1972 than in 1976. And so they were. But the elements of an electoral inversion so clearly seen in 1968 and 1972 were not wiped away in the Carter-Ford contest. There was a fairly steady increase in the Ford proportion with movement up the socioeconomic ladder, however socioeconomic status is measured. But there was no return to the class voting of the New Deal years. Louis Harris found that grade school- educated whites were about 14 percentage points more Democratic in 1976 than were their college-trained counterparts.

Whites from families earning less than $5,000 a year were approximately 16 points more for Carter than were whites with annual family incomes of $15,000 and higher. Data made available from the Election Day "exit" polls of the New York Times, CBS News, and NBC News show similar distributions. In the 1930s and 1940s, by way of contrast, high-status and low-status whites were separated by between 30 and 40 percentage points in presidential preference.

Lower-status cohorts were relatively more Democratic, compared with their upper-status counterparts, in the contest between Gerald Ford and Jimmy Carter than in either of the two preceding elections. But the Democratic contender in 1976 did not regain the big relative margin among lower-class voters enjoyed by his Democratic counterparts in the New Deal years, nor that which Kennedy achieved as late as 1960. Carter ran a campaign which minimized rather than heightened the "social issue" concerns of lower-status whites. And it should be noted that actual social conditions in 1976 made it far easier for a Democratic nominee to accomplish this than it had been in 1968 and 1972. Still, the altered meaning of liberalism--and hence of the national Democrats as the liberal party--together with the changed social position of the white working classes and the professional/managerial groups, had precluded a return to the class voting patterns of the New Deal epoch.

College-trained professional and managerial people, especially the younger age groups, remain consistently more liberal than the noncollege blue-collar groups on a wide range of social and cultural issues. On matters of economic policy, however, they are often more conservative, although much less decidedly so than during the New Deal. The Democratic party came, in the 1960s and 1970s, to be associated with both the new social liberalism and the old economic liberalism, and as a consequence it evoked contradictory responses within both the upper and the lower socioeconomic categories. The decline in class distinctiveness in voting, and the tendencies toward inversion, came as a result--among white Americans. Black Americans must be treated separately, because their electoral choices have not been accounted for by the class/ideology dynamics we have been describing [3]


Since the 1980 presidential balloting, a lively debate has ensuedover what the election results actually mean and what their long-term consequences are likely to be. Some have argued that a marked ideological change has occurred in the United States inrecent years--the populace has "swung to the Right"and thatRonald Reagan and the Republicans are building what is likely to be a lasting new majority on the more conservative public mood. Others maintain that the election was simply the rejection of an ineffective president who had to confront some intractable problems, and that at most the GOP won "an opportunity" to show that it could govern successfully.

That this argument has centered on whether a new realignment burst upon the U.S. political scene in 1980 seems to me most unfortunate because it is so untheoretical. As noted, new alignments have been emerging in response to broad social changes ever since the New Deal, and especially since 1960. The 1980 balloting signalled the continuation and the elaboration ofthese changes, not some sudden new departure. And it has been evident over the 1970s that the primary direction of these shifts is toward dealignment, not realignment.

A realignment encompasses the movement of large numbers of voters across party lines, establishing a stable new majority coalition. In a dealignment, by way of contrast, old coalitional ties are disrupted, but this happens without some stable new configuration taking shape. Voters move away from parties altogether; loyalties to the parties--and to their candidates and programs--weaken, and more and more of the electorate "comes up for grabs" each election.

Over the 1970s an impressive body of theory and data developed which supports the argument that the present era in American electoral politics is one of dealignment and party decay, which specifies the sources of this development, and which identifies its implications for American political life. The progress of dealignment has been giving recent electioneering its distinctive cast.

During the late 1960s and early 1970s one began noting, with increasing frequency, signs of a vast transformation of the political parties and of the nature of voters' attachments to them. For one thing, the national party organizations, never notably robust, experienced a major decline and loss of functions. Within the Congress and in the conduct of presidential nominations, political individualism triumphed over institutional and collective requirements. I have described these developments in some detail elsewhere (Ladd 1979, 1981c).

A weakening in public attachment to the parties has contributed to their institutional decline as much as it has resulted from that decline. Increasing numbers of voters think of themselves as independents, and regularly cross party lines with abandon in election contests (Ladd 1978, pp. 320 - 33). The proportion of the public lacking confidence in both of the major parties has grown substantially (Ladd 1981a, pp. 3 - 7). The electorate is less shaped and located by stable partisan loyalties now than at any time in the last century.

An electorate thus "cut loose" is potentially much more volatile than one held in place by the anchors of party. This potential for volatility that party dealignment creates was amply realized over the 1980 presidential campaign, but it had been apparent for at least the preceding decade (see Ladd 1 1981a 0 1 1981b ).


Writing in 1971, Samuel Lubell argued that the Republican coalition showed little prospect of becoming majoritarian and that "there does not seem to be much chance of reestablishing the New Deal coalition in its old form" (Lubell 1971, p. 278). He saw party loyalties becoming ever more lightly held, and party fading into the background before what he called "total elections"those in which presidents utilize their power in economic and foreign affairs in an attempt to orchestrate electoral results. When things go wrong in this setting, the president is blamed massively and is banished from office. Thus, the United States had arrived, Lubell argued, at an age of plebiscitary presidencies. These developments, contributing to an inability to establish any long range resolution of conflicts dividing the country, had produced "a new alignment of two incomplete, narrow-based coalitions polarized against each other," with the bulk of the populace dealigned and responding afresh every four years to the presidential "referendum". Lubell was right.

The changes occurring in the parties and in the public's loyalties to them over the 1960s and 1970s thus were not building to a "critical realignment" of the sort that took place in the late 1920s and 1930s. The changes were momentous enough, and they brought about the dissolution of the New Deal party system, but they were not yielding a coherent new majority. Old attachments were being disrupted without, for the most part, new ones being built in their place. The full impact of electoral dealignment was to be felt in the volatile, party-free voting of 1980.